Loading datasets...
Task: ASR
Release Date: 3/20/2026
Format: MP3
Size: 202.32 MB
Share
A collection of spontaneous responses to questions in Gheg Albanian (aln).
Restrictions/Special Constraints
None provided.
Forbidden Usage
It is forbidden to attempt to determine the identity of speakers in the Common Voice datasets. It is forbidden to re-host or re-share this dataset.
Intended Use
This dataset is intended to be used for training and evaluating automatic speech recognition (ASR) models. It may also be used for applications relating to computer-aided language learning (CALL) and language or heritage revitalisation.
aln)This datasheet is for sps-corpus-3.0-2026-03-09 of the Mozilla Common Voice Spontaneous Speech dataset for Gheg Albanian [aln - aln]. The dataset contains 1654 clips representing 10.86 hours of recorded speech (10.49 hours validated) from 14 speakers.
Gheg Albanian (/ɡɛɡ/) is an Albanian dialect group spoken in northern and central Albania, Kosovo, northwestern North Macedonia, southeastern Montenegro, southern Serbia and southwestern Croatia. Identified by the ISO 639-3 code aln, it belongs to the Indo-European language family. There are estimated to be around four million speakers. There is no widely accepted writing system.
The dataset clips are categorised by transcription status and training-set assignment. The following tables summarise the distribution.
| Bucket | Clips | % |
|---|---|---|
| Transcribed & Validated | 1,516 | 91.7% |
| Transcribed & Pending | 0 | 0.0% |
| Not transcribed | 138 | 8.3% |
| Bucket | Clips | % |
|---|---|---|
| Train | 934 | 56.5% |
| Dev | 300 | 18.1% |
| Test | 282 | 17.0% |
| Unassigned | 138 | 8.3% |
Training split coverage: 1,516 of 1,516 transcribed & validated clips (100.0%)
The transcription of the entire dataset was done by a single individual, using the following guidelines:
In general, the standard Albanian writing system was used, with the following exceptions:
<Ε> was used in place of standard Albanian <ë> /ǝ/
<C> was used in place of standard Albanian <ç> /tš/
<gh> was used to represent /γ/
<hj> was used to represent /ç/
Prompts: 144
Duration: 39113172[ms]
Avg. Transcription Len: 198
Avg. Duration: 23.63[s]
Valid Duration: 37793.74[s]
Total hours: 10.86[h]
Valid hours: 10.5[h]
| Bucket | Clips | % |
|---|---|---|
| Validated | 1,516 | 100.0% |
| Pending | 0 | 0.0% |
| Edited | 103 | 6.8% |
There follows a randomly selected sample of questions used in the corpus.
Njerzit që kan mosh të ndryshme si e përdorin teknologjin?
Pse menon se qielli asht blu?
Masi ke ba familje ça ke ba?
A din nonj nofk në shqip?
A ke shku ndonjëher me avion?
There follows a randomly selected sample of transcribed responses from the corpus.
*Pwr zotin, munohena me i msu kalamajt me kalamajt tash a kupton... jena tu u munu me i msu kalamajt me jo msusja jo tjetra, e marrim msusen e Cojm kur vjen msusja, vjen ndonjher rrall dhe shkon msusja, a kupton, e mson msusja, n shkoll shkojn e msim, munohena me ta ke shpia me msu ke shpia mir, po sa... *
*Pwr pushime mw pwlqen me ik nwpwr ishuj ktu n Greqi, mw pwlqen mw shum me shku nw Cdo vjet edhe nga njw ishull tw ndryshwm, edhe pwr me i vizitu se si jan edhe pwr me i pa nga afwr, edhe ngaqw mu m pwlqen shum plazhi, mw shum plazhi, domethwn a don me zgjedh me shku n mal apo don me hik n plazh, un gjithmon zgjedh plazhin, e edhe kombinimi i tw dyjave ashtw prap ma i veCant edhe ma i bukur edhe do t ishte njw mrekulli e vwrtet. *
Me bo kamarjerin.
Mana, po, un di me i ra gitar, un di me i ra edhe Ciftelis, e jam shum i mir n Cifteli, edhe shishto, pwr zotin, po.
Cila lloj swmurje je tu kalu rin ti? Ca lloj swmurje jesh ka kaluar ti? Lloj smundje kam pas kalu, grip kam pas kalu, a m kupton, grip, grip mw ka pas kap para nji jave, gripi, gripi them na, po, grip, na ka kap, kena pi Caj, po, Caj, po, po.
Each row of a tsv file represents a single audio clip, and contains the following information:
client_id - hashed UUID of a given user
audio_id - numeric id for audio file
audio_file - audio file name
duration_ms - duration of audio in milliseconds
prompt_id - numeric id for prompt
prompt - question for user
transcription - transcription of the audio response
votes - number of people that who approved a given transcript
age - age of the speaker1
gender - gender of the speaker1
language - language name
split - for data modelling, which subset of the data does this clip pertain to
char_per_sec - how many characters of transcription per second of audio
quality_tags - some automated assessment of the transcription--audio pair, separated by |
transcription-length - character per second under 3 characters per second
speech-rate - characters per second over 30 characters per second
short-audio - audio length under 2 seconds
long-audio - audio length over 5 minutes
Antonios Dimakis <a.dimakis@athenarc.gr>
This dataset was partially funded by the Open Multilingual Speech Fund managed by Mozilla Common Voice.
This dataset is released under the Creative Commons Zero (CC-0) licence. By downloading this data you agree to not determine the identity of speakers in the dataset.
For a full list of age, gender, and accent options, see the demographics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2